Cameras in the Courtroom: An Experiment on the Public’s Views

1. Access to the Court

To understand the debate over cameras in the Court, one must know something about the history of oral arguments at the Supreme Court. The purpose of these proceedings, and access to them, have evolved over time. The early Court relied primarily on oral arguments (not briefs) to obtain information (Johnson, 2004; Stern, 1989) and imposed few rules on the length of attorneys’ arguments. As a result, proceedings often lasted days (Rehnquist, 2001, p. 242). People in and around Washington regularly attended oral argument (White, 1988). The courtroom often bulged at the seams with viewers, particularly since the young Washington D.C. was yet a “dismal place” littered with “small miserable huts” and little entertainment (Rotunda, 2018, pp. 124–125). As an example, when the Court heard McCulloch v. Maryland (1819), Justice Joseph Story wrote to a friend: “We have had a crowded audience of ladies and gentlemen…The hall was full almost to suffocation, and many went away for want of room” (Rotunda, 2018, p. 414).

It was predominantly those near Washington who could travel to oral argument. “In New England in 1790, vehicles were few, roads were generally rutted and rudimentary, and traveling any distance was both slow and difficult…” (Historical Background, n.d). Even the strongly Jeffersonian and states’ rights Secretary of Treasury Albert Gallatin recognized the tremendous difficulties of travel when he proposed a $20 million national transportation plan (Gallatin, 1807). The result of these conditions was that access to the High Court was limited.

Years later, the Court expanded access to oral arguments. It began recording audio of the proceedings during the 1955 term. Still, the justices did not release those audio recordings until after the completion of each term – and then only to the National Archives. So, people had to wait up to nine months to hear the cases and only those willing and able to seek out the reel-to-reel tapes could hear the justices in action. In 1993, the Court began to make these recordings accessible earlier than the end of the term. Beginning in 2000, the Court began to release same-day audio recordings of proceedings (e.g., Bush v. Gore). Late in the 2019 term – during COVID – the Court offered live-streaming audio of its cases (Ringsmuth et al., 2023). That practice continues today.

Though the Court has expanded auditory access to oral argument, visual access remains limited. The Court sets aside 50 seats for the public to watch an entire oral argument, but those spots are difficult to obtain. (The Court also holds roughly 25 seats “in the back of the courtroom for members of the public who wait in the ‘three-minute line’” (Howe, 2020) but it seems unfulfilling to observe the Court for a mere three to 5 minutes.) And so, people who wish to observe the justices in person often go to great lengths to obtain a seat. For a while, professional line-standers offered their services to people who sought to attend an argument (Bart, 2020). In December 2024, the Court replaced those physical lines with an online lottery system (Howe, 2024). But that reform still does not address the capacity issue. Given the interest in Court cases and the limited physical access, many people call for the justices to allow cameras during oral arguments so that the public can observe its proceedings. ¹

Camera supporters make a variety of claims beyond the capacity issue referenced above. First, they argue cameras might make the Court more transparent and accountable to the public. Similarly, cameras might educate and inform the public about cases. They also could show people how the otherwise opaque Court operates. Further, in a visually dominant society, people might be more engaged with video than they are with audio. Putting cameras in the Supreme Court could also enhance its legitimacy. Justice Elena Kagan represented this position when she stated: “[Cameras] would allow the public to see an institution working thoughtfully and deliberately and very much trying to get the right answers, all of us together” (Wolf, 2019). Research indicates that “to know the court is to love the court” (Gibson & Caldeira, 2009). Under the right conditions, cameras might enhance public support (Bybee, 2010).

Camera opponents take the opposite view. They argue that cameras will not make the Court more transparent because the justices’ actual decision making would continue to take place in its private conferences (Black & Johnson, 2018) and during its opinion writing process (Maltzman et al., 2000). What is more, the public is not supposed to influence the Court. It is a legal institution designed to protect legal rights against majoritarian attacks. Its ultimate accountability is to the U.S. Constitution, not to the public. Further, they believe the media may use video footage to misrepresent Court proceedings, which could harm judicial legitimacy. Former Justice Antonin Scalia expressed this precise apprehension:

…I am sure it will miseducate the American people, not educate…what most of the American people would see would be 30-second, 15-second takeouts from our argument, and those takeouts would not be characteristic of what we do. They would be uncharacteristic (C-SPAN, 2012). ²

To be clear, not all justices are opposed to cameras. Some have indicated they would entertain cameras but only if their proponents can clearly establish the benefits. They want data on the consequences of cameras. As Justice Stephen Breyer stated: “[A] decision of this issue [cameras in the courtroom]…which carries with it threats to that institution as well as benefits, should be decided after really pretty serious research and study…” (Kennedy, O’Connor and Breyer 2006, pp. 85–86).

Surprisingly, neither camera supporters nor opponents have done much to answer Breyer’s call for data. The most common empirical approach to the topic has simply sought to poll the public on the matter. As of March 2022, 65% of respondents agreed with the statement “The U.S. Supreme Court should allow television coverage of its oral arguments” (C-SPAN, 2022, p. 14). ³

The behavioral side of the question – whether cameras change what judges actually do – has only recently received empirical attention. Kaufman et al. (2026) leverage the quasi-random rollout of video recording in the Ninth Circuit to test what they term the “performative judging hypothesis,” using AI-assisted diarization of oral argument audio. They find suggestive evidence that judges speak more and interrupt attorneys more frequently when proceedings are video recorded, though their results are mixed and model dependent. Their study examines how cameras might affect those inside the courtroom. Whether cameras affect the views of those outside it – the public – requires a different kind of evidence.

Only a handful of rigorous studies examine cameras at the Supreme Court. Reviewing media-effects research, Carter (2012) concludes that the educational and civic-participation benefits of broadcasts would outweigh the potential negatives, while acknowledging that more direct empirical work is needed. He specifically calls for simulation-based studies using oral argument footage from other courts as a means of generating empirical evidence in the absence of actual Supreme Court video – an approach our experimental design adopts.

Building on the kind of approach Carter recommends, our own prior work contributes to this still-developing literature. In a previous study, we examined whether watching content from oral argument in state court proceedings affected respondents in ways different than listening to it (Black et al., 2023a). ⁴ We found that static camera angles did not appear to influence the public’s support for the Court but that dynamic camera angles might do so. Respondents who watched a neutral exchange (via a dynamic angle) between a judge and lawyer showed an increase in court support but respondents who watched a contentious exchange (via a dynamic angle) showed a decrease in court support. We also found that the presence of judicial symbols could reduce the otherwise negative effects of observing a contentious exchange (Black et al., 2023a).

Building upon these initial results, we also evaluated whether watching versus listening to this content influenced the amount of trust that someone had in courts “to make decisions that are right for the country as a whole” (Black et al., 2024). Relative to a control group, who neither saw nor heard content, exposure to any content tended to result in a substantively small but statistically significant increase in trust (id. at 345). There was no evidence that cameras increased trust in courts under any conditions and some evidence that, under unfavorable conditions, they might decrease trust (id. at 347).

We also found that people’s views of particular judges can change after they watch oral argument (Black et al., 2023b). Respondents who watched a neutral exchange between a judge and attorney wound up supporting the judge more than those who merely listened to that neutral exchange. Additionally, respondents thought no less (nor more) of a judge when they watched him in a contentious exchange with an attorney versus listening to that exact same contentious exchange. But respondents supported the judge significantly less when they watched him engaged in a contentious exchange versus watching him in a neutral exchange.

These studies, however, did not examine the concepts that are of interest in this paper. They did not analyze people’s broader views on whether (and when) courts should allow cameras in their courtrooms, whether oral argument is perceived to benefit society, people’s perspective on whether it is important to use oral argument versus briefs alone, and whether judges should avoid frequent disagreements with attorneys and other judges.

We believe the first two of these questions directly address how people would react to cameras in ways heretofore unexamined. Would people more likely support cameras once they are able to observe oral argument through their lens? It could be that people care little about cameras, and seeing an oral argument might make them less interested in seeing one again – or at least make them less interested in having cameras in the Court. Similarly, it could be that, once having seen oral arguments, people believe they are not particularly beneficial to society. If so, then the intensity with which people hold views over cameras in the Court may be considerably less than camera supporters believe. Indeed, previous work on possible court reforms suggests that, as soon as people discovered there were opportunity costs to seeking retirement ages or term limits for federal judges, they became much less interested in these changes (Black et al., 2023c, 2024). Perhaps after people actually observe oral argument through video provided by cameras placed in the courtroom, they become less enamored with the proceedings.

The third question goes directly to the heart of whether people support oral argument versus briefs alone. In other words, do people think these proceedings offer something above the written briefs? Admittedly, this question asks much of our respondents, many of whom may not know the difference between oral argument and brief writing. Still, it provides a datapoint on that debate.

Our fourth question is useful because it gets at what justices likely find most useful about oral argument – the ability to follow up and press attorneys on their arguments. Johnson (2004, p. 5) argues that justices often “seek new information during these proceedings” to help them reach decisions. Oftentimes, to obtain such information, justices must interrupt attorneys or disagree with them to push their arguments. The exchange can be uncomfortable though beneficial. For those uninitiated with oral argument, the exchange may look overbearing and downright mean (of course, it could be all these things). And so, if a respondent observes oral argument and walks away from it thinking that judges should avoid disagreements with attorneys and other judges, they might seek to kill the very part of oral argument justices find useful.

2. Survey Experiment

To provide more perspective on the issue of cameras in the nation’s High Court, we use the experimental data collected in our initial research (Black et al., 2023a, 2023b, 2024). More specifically, this research employed Lucid Theorem to assess the effects of observing portions of an oral argument. ⁵ The experiment utilized 1,475 respondents and a 2 × 2 × 2 plus control design. Individuals were randomly assigned to one of nine conditions. The treatments consisted of a single 50–60 second clip of one of two state supreme court oral argument exchanges between an attorney and a justice. ⁶ The approach was a between-subjects design with post-treatment measurement of the outcome variables.

Though scholars cannot examine videos of U.S. Supreme Court oral arguments, we opted for a realistic alternative: state supreme court videos. ⁷ State supreme courts are courts of last resort for matters of state law. They often share similarities with the U.S. Supreme Court. Justices wear black robes and sit upon elevated daises. A member of the public would be unlikely to distinguish the interior courtroom of a state supreme court from the U.S. Supreme Court. To be sure, we did not deceive respondents into thinking they were watching the U.S. Supreme Court. We simply referenced a “court” and then let respondents observe.

Respondents watched or listened to clips from the Minnesota and Indiana supreme courts. ⁸ We selected these two courts for a number of reasons. First, we sought to mitigate possible confounders that might influence respondents’ reactions, such as their preferences for or against certain accents (Amira et al., 2018). Because the two accents in the sample were similar, respondent differences could not be attributed to accent preferences. Second, we needed quality videos. Few states provided video footage, let alone consistently decent quality footage. After reviewing available state court videos, we determined those two were the best. They also provided important variation for one of our treatments (camera angle) on which we elaborate more below.

Respondents read a few sentences of background material about the case stimuli they would see or hear. They were then exposed to our treatments, which manipulated the clip’s modality: respondents either watched a video clip (with sound) or listened to the audio without video. Those assigned to the control group neither viewed nor listened to an oral argument clip. ⁹

The design also manipulated the contentiousness of the exchange between the judge and attorney. ¹⁰ An exchange was “contentious” if it was quarrelsome or disputatious. In the contentious clip, a justice aggressively questioned the attorney, interrupted him, raised his voice, and appeared visibly impatient. The neutral clip featured the same attorney, the same justice, in the same case, and on the same general topic, but showed them engaged in a different set of neutral exchanges that lacked these contentious behaviors. ¹¹

Finally, we manipulated the angle of the camera. As Figure 1 shows, the Minnesota Supreme Court – panel (a) – employed a static wide-angle shot from a distance that captured all the justices as well as a side view of the attorney. The viewer could not easily see facial expressions but was able to observe all the actors. Conversely, the Indiana Supreme Court – panel (b) – employed multiple cameras that focused on whomever spoke at a particular moment. These cameras allowed viewers to see the speakers’ faces and their expressions. The flip side is that it did not permit the viewer to see the full court all at once.OPEN IN VIEWER

Our outcome measures ascertained whether people who watch oral argument will have different attitudes about the proceedings relative to people who listen to them or do not observe them. To that end, we leveraged four post-treatment statements. Using a five-point Likert scale (i.e., strongly agree, somewhat agree, neither agree nor disagree, somewhat disagree, strongly disagree), we measured respondent attitudes towards statements about courts and judges. Two of these statements focused on cameras or observing judges at work, and two sought to probe the value of oral argument.

The first dependent variable was a respondent’s agreement with the statement that “All courts should allow cameras in their courtrooms so that anyone who wants to watch oral argument can do so.” The second dependent variable measured a respondent’s support for the statement that “Allowing the public to observe judges and courts doing their work offers little benefit to society.” While the first dependent variable mentioned cameras explicitly, the second did not. In contrast to the first measure, this question was also phrased in the negative to protect against acquiescence bias.

The third dependent variable was a respondent’s agreement with the statement that “It is important for attorneys and judges to have oral argument rather than making decisions based on written arguments only.” One of the critiques of oral arguments is that they do not add much to the justices’ decision-making process. While research has shown the opposite (e.g., Black et al., 2012; Johnson, 2004), it remains to be seen how exposure to oral argument influences the public’s attitudes toward them.

The fourth dependent variable measured agreement with the following: “During oral argument, judges should avoid frequent disagreements with the attorneys and other judges.” Judges and attorneys occasionally clash. This is all part of the appellate litigation process. Sometimes, judges ask tough questions to elicit answers about the logical limits of an attorney’s position. Attorneys need to advocate zealously for their client and defend their arguments. These exchanges can seem combative and harsh. While attorneys and judges (usually) see value in those exchanges, members of the public might find them uncomfortable or off-putting.

3. Methods

To make the results easily comparable, we rescaled each of our four dependent variables from 0 to 1 such that 1 indicates more support for the pro-cameras/access position associated with each of our four dependent variables. ¹² The appendix contains their descriptive statistics and graphically displays their respective distributions. Because the outcome measures tap conceptually distinct dimensions, we separately examined each outcome measure and, for each measure, we evaluated comparisons via two-sample difference-in-means tests. ¹³ To understand what follows, it is necessary to explain how our analysis builded in an orderly progression.

Our approach began with the most basic question: were there unconditional differences between respondents who experienced any oral argument content versus those in the control group? We then evaluated the difference between any audio content versus any video content. This allowed us to examine whether there was a medium effect when pooling across the non-control conditions. This step was important because audio already exists – we were interested in learning about the potential impact of adding video.

If we found a medium effect, then we drilled down deeper to ascertain what might be driving those results. We did so by making comparisons that held constant the video conditions so that we could examine whether the content (neutral versus contentious), camera angle (static versus dynamic), or some interaction between content and angle, were responsible.

In the results section be, each comparison we present plots the average differences from two-tailed difference of means tests between two respective conditions and then highlights the significance of our results. When considering a treatment’s effect, we report the two-tailed p-values in each figure. ¹⁴ Given the exploratory nature of our analysis, our discussion below focuses on relationships whose p-values are less than or equal to 0.10, consistent with the view that significance thresholds should be justified by research context rather than fixed by convention (Lakens et al., 2018).

4. Results

5. Conclusion

There is a growing debate over whether to allow cameras in courtrooms, particularly at the U.S. Supreme Court. This debate, however, tends to occur without the benefit of systematic data to help reach a conclusion. Few studies empirically investigate how people react to observing videos of courtroom proceedings. We have tried to do so here.

A potentially useful way to make sense of these mixed results is to group our four outcomes along two dimensions. Some items concern judicial behavior – whether courts should allow cameras and whether judges should avoid open disagreement – while others concern the judicial process itself. ²¹ Viewed this way, a pattern emerges. The two behavior-oriented attitudes both moved in response to exposure, and both moved in the direction of greater openness: respondents who saw oral argument became more supportive of cameras and more accepting of the on-the-record disagreement that characterizes real appellate proceedings. The two process-oriented attitudes, by contrast, were far less responsive. Attitudes about the value of oral argument relative to briefs did not move at all, and attitudes about the societal benefit of public observation moved only in response to video – and ultimately in the opposite, less favorable direction.

We offer one interpretation of this divergence, while acknowledging that our exploratory design cannot adjudicate among competing explanations. Attitudes about how judges should behave may be comparatively easy to update because most members of the public have little prior exposure to what a real oral argument looks like; seeing one provides new, concrete information that can revise expectations about whether courtroom disagreement is normal or troubling. Attitudes about the process (i.e., whether observation is socially beneficial, whether oral argument matters) may rest on more settled priors about the role of courts that a brief exposure is less likely to disturb. This would explain both why the process items were stickier and why the one process attitude that did move (the benefit of observation) moved only under the richer video treatment, which conveys more of the texture of a proceeding than audio alone.

The persistent ambivalence about whether public observation benefits society deserves particular emphasis, as it is both substantively striking and, to our knowledge, underexplored. Even after exposure, a substantial share of respondents remained unconvinced that watching judges work offers much social value – and video exposure, if anything, deepened that skepticism. This sits in tension with a common assumption among camera advocates: that if the public could simply see the Court at work, support for transparency would follow. Our results suggest the relationship is more complicated, and that exposure may inform attitudes about judicial conduct more readily than it builds the case for transparency as an end in itself. In that sense, our findings echo Strauber’s (2016) concern that confident predictions about how the public will respond to cameras are not well-supported by the available evidence – exposure shifts some attitudes but not others, and not always in the direction camera advocates might expect.

The movement we do observe (i.e., toward greater support for cameras and greater acceptance of courtroom disagreement) is particularly notable in light of recent evidence that cameras may change the very behavior respondents are evaluating: Kaufman et al. (2026) find suggestive evidence that circuit judges speak more and interrupt attorneys more often when video recorded. If cameras simultaneously alter judicial conduct and the public’s tolerance for that conduct, the net effect on perceptions of the Court is even harder to forecast, thereby reinforcing our view that adoption should await further study.

These patterns add texture to the mixed results without resolving them into a single recommendation. For us, this underscores the inherent complexity of the topic, and, by our reckoning, there is not a simple answer to whether the U.S. Supreme Court should allow cameras to cover its oral arguments. In that sense, our most important conclusion is to emphasize, highlight, and underscore the quotation we presented earlier from former Justice Breyer – namely that a decision of this magnitude is one that should only be taken after “pretty serious research and study.”

Our efforts, while serious, should be understood as early-stage empirical work subject to several limitations. At the top of the list is the fact that the present debate is ultimately about a very specific and very unique court – the U.S. Supreme Court. Since cameras are presently not allowed, we used materials from other courts for our experimental stimuli. While real world content has the advantage of being externally valid, it limited the extent to which we could strictly manipulate several aspects of our stimuli. Among other things, the videos differed not only in camera angle and the contentiousness of the exchange between the judge and the attorney, but also in video quality, the exact content of the exchange, whether the judge’s name was displayed, and a host of other features that may have – consciously or unconsciously – affected participants’ reactions. Future efforts should push towards achieving better approximations of the Supreme Court in their experimental materials with an enhanced ability to manipulate all aspects of the stimuli.

Efforts to open the Supreme Court – and other judicial bodies – to cameras are unlikely to end anytime soon. It is imperative that there be an objective, empirical assessment of the claims made by both sides of this weighty debate. In so doing, researchers will be able to help courts assess both the possible strengths and shortcomings that such changes to the status quo might create.

Supplemental Material